View sample language development in childhood research paper. Browse research paper examples for more inspiration. If you need a psychology research paper written according to all the academic standards, you can always turn to our experienced writers for help. This is how your paper can get an A! Feel free to contact our writing service for professional assistance. We offer high-quality assignments for reasonable rates.
Somehow in the span of just a few years, newborn infants who neither speak nor understand any language become young children who comment, question, and express their ideas in the language of their community. The transition from the stage of the prelinguistic infant to the linguistically competent 4-year-old follows a predictable developmental course. First, newborns’ cries give way to coos and babbles. Then, infants who coo and babble start to show signs of comprehension such as turning when they hear their name. Infants then become toddlers who say bye-bye and all gone and start to label the people and objects in their environment. As their vocabularies continue to grow, children start to combine words. Children’s first word combinations, such as all gone juice and read me, are short and lack parts found in adults’ sentences. Gradually, children’s immature sentences are replaced by longer and more adult-like sentences. As children master language, they also become masters at using language to serve their needs. One-year-olds, who can only point and fuss to request something, become 2-year-olds who say please; later they become 4-year-olds capable of the linguistic and communicative sophistication of the child who excused himself from a boring experiment by saying, “My mother says I have to go home now” (Keller-Cohen, January 1978, personal communication).
This course that language development follows is the result of concurrent processes of development in the several domains that together constitute adult knowledge of language. In acquiring language, children master a system for combining sounds into units of meaning (phonology) and a system for combining units of meaning into well-formed words and sentences (morphology and syntax). Together these systems constitute the grammar of language. In addition to grammar, a speaker-hearer of a language has acquired a lexicon—a repository of words with their meanings and grammatical categories. Normally, speaker-hearers of a language use that language to communicate. Thus, adult-like language competence includes pragmatic or communicative competence as well as knowledge of a grammar and lexicon. The goal of this research paper is to describe the current state of the scientific effort to explain how these changes take place and how children end up knowing a language. The content of this research paper is not a description of language development; those are readily available elsewhere (e.g., Hoff, 2001). Rather, this research paper provides a description of the field that takes language development as its topic of study.
This description focuses on research that seeks to describe the nature of the mental capacity that underlies the human ability to acquire language. The fact that language acquisition occurs in a range of social and cultural environments is relevant to constructing such a description, and research aimed at identifying the necessary social conditions for language acquisition to occur is reviewed. A more complete consideration of how culture influences language development falls under the heading of language socialization, that is, the process by which children come to use language in the manner of their social or cultural group. That topic is touched on only briefly in this research paper; in-depth treatments can be found in Ochs and Schieffelin (1979), Schieffelin and Ochs (1986), and Slobin, Gerhardt, Kyratzis, and Guo (1996).
The Question to Which Research on Language Development Is Addressed
The study of language development is marked by serious disagreement with respect to both what the correct explanation of language development will look like and how best to discover that explanation. There is, however, an abstract level at which all researchers in the field are trying to answer the same question: What is the nature of the human capacity to acquire language? This question can be conceptualized in the following manner: The human capacity for language is a device residing in the human brain that takes as its input certain information from the environment and produces as its output the ability to speak and understand a language. Everything that is part of adults’ knowledge of language (i.e., the output of the device) must either be in the input, be in the internal device, or somehow result from the way the device operates on the input it receives (see Figure 7.1).
Noam Chomsky (1965) termed this capacity the Language Acquisition Device (LAD), and this particular way of posing the question still tends to be made explicit only by those taking a generative grammar approach to the study of language acquisition. However, this conceptualization makes clear two criteria that must apply to all candidate explanations of how children learn to talk. To wit, any proposed account of the language acquisition process must be consistent with two sets of facts: (a) the input that children receive and (b) the competence that they acquire. It is obvious that language acquisition makes use of input because children quite reliably acquire the particular language to which they are exposed. Yet it is not at all clear—and a matter of great dispute—whether input provides sufficient information to explain how children end up knowing what all adult speakers know. It is also clear that what children acquire is knowledge of a productive system because adults can and do understand and produce sentences that they have never heard before. Just how best to characterize that system is also a topic of dispute. As we review the current work in the study of language development, we will see the sometimes stark differences in the views of input and ultimate competence that guide research in this area. These differing views crucially affect the nature of the debate regarding how language is acquired because the problem of explaining acquisition is different depending on the nature of the input and the nature of what is acquired.
Theoretical Approaches to Discovering How Language Is Acquired
Current research on language development can be usefully organized as being motivated by four different premises regarding the nature of the LAD and the language development it produces. One such premise is that the human capacity for language is best understood as a biological phenomenon and language development as a biological process. This leads to research investigating the degree to which language and language development share the hallmark features of other biological processes, such as universality and heritability, and to the study of the anatomical structures and physiological processes that underlie language development. A second premise, which guides other research in the field, is that language acquisition is best understood and studied as a linguistic phenomenon. On this view, the LAD is essentially a linguistic device. This approach starts with a description of the linguistic competence of adults and seeks an account of the acquisition process that is adequate to explain how that end point is reached. This approach is often referred to as the learnability approach because its focus is on explaining how it is that language is learnable. The biological and linguistic approaches share the view that language is innate. For the biological approach,innatenessispartofthepackage—somethingthatis part of human biology is, by definition, innate. For the linguistic approach, innateness is a quickly reached conclusion from the following line of reasoning: The complex, abstract system that is adult linguistic competence simply could not be arrived at by any general learning mechanism operating over the input that children receive. The solution to this problem, known as the logical problem of language acquisition (C. L. Baker & McCarthy, 1981), is to propose that substantial language-specific knowledge and language-specific learning mechanisms are part of the equipment that children bring to the language-learning task. The charge for research is to describe that innate knowledge (Crain, 1991).
There are alternative approaches that resist attributing innate linguistic knowledge to the child and that seek a description of how children could learn language from experience using learning mechanisms that are not specific to language. The chasm between the learnability approach and the developmental approach is such that even labeling the field carries implications of allegiance to one view or the other. The term “language acquistion” is most associated with the learnability approach; the terms “language development” and “child language” are most associated with the developmental approach. Research within this developmental approach, as it is termed, begins not with the end point of development, but rather with the starting point—the child’s demonstrated competencies. Such work seeks to explain not the fact that language is acquired but the developmental course that language learning follows. Within the developmental approach, two different lines of argument and research exist. One starts from the premise that language is essentially a social phenomenon and that language development is a social process. Such research focuses on social aspects of interaction as the experience relevant to language acquisition and on the social-cognitive abilities of the child as the relevant learning capacities. The other starts from the premise that language acquisition is essentially an asocial learning problem that children solve in the same way they solve other learning problems. Research in this vein seeks an account of how language might be learned by the child’s application of domain-general cognitive processes to information available in input.
The goal of this research paper is to present each of these four approaches: the biological, the linguistic, the social, and the domain-general cognitive. None of these presentations is comprehensive. Rather, for each approach the aim is to illustrate the nature of the research and theoretical argument it generates and to evaluate its contribution to explaining how children learn to talk.
Language Development as a Biological Process
If we begin with the premise that the LAD is a biological entity and its operation a biological process, we are led to investigate the degree to which language acquisition shares hallmark features of other biological processes and to investigate how the anatomical structures and physiological processes that accomplish language acquisition actually do their work. The hallmark features of biologically based characteristics include species universality and species specificity, an invariant course of development that is robust over varying environmental circumstances, a critical period for development, heritability, and an adaptive function that explains its evolution as a characteristic of the species. To the extent that human language and language acquisition meet these criteria, language and the capacity for language acquisition would seem to be part of human biology. That is to say, language would appear to be innate.
The Species Universality and Species Specificity of Language
All humans have language, and no other species has a communication system that shares all the features of human language. In addition, in the absence of a language to learn, humans will create one. For example, it has been widely observed that deaf children in hearing families invent systems of signs with which to communicate. These systems have the equivalent of syntax and morphology and a lexicon in which different words belong to different grammatical categories, and the system as a whole is used for the same sort of purposes as are established languages, thus demonstrating the basic features of all human languages. Because children invented rather than learned these home sign systems, it is argued that these features reflect components of language that are built into the human mind (Goldin-Meadow, 1997).
Other evidence similarly suggests that the necessary and sufficient ingredients for language creation are the opportunity for communication with others and a human mind—particularly a child’s mind. Although socially isolated children do not invent languages (Shatz, 1994), whenever people come together, language emerges. Pidgins are languages that emerge when circumstance puts together people who share no common language. Pidgins tend to be morphologically simple languages, lacking markers of subject-verb agreement, tense, and so on. With time, and with the birth of children who acquire these pidgins as their native language, the pidgins evolve into creoles, which have more elaborate grammatical morphology. It has been argued—although not universally accepted (e.g., Jourdan, 1991)—that children play a crucial role in the process of creolization and that the structural similarities among creole languages with independent origins suggest the work of the human capacity that is normally put to language acquisition, but which in the absence of a full-blown language to learn, reveals itself as the human capacity for language creation (Bickerton, 1984, 1988).
More recent evidence for the human capacity—and particularly the human child’s capacity—to create language comes from the study of Nicaraguan Sign Language (NSL). This language has emerged in just the last 25 years, following the opening of the first public schools for the deaf in Nicaragua in 1978. When they entered the school, the deaf children typically had only their own idiosyncratic home sign systems and no shared language, but in the school setting a new sign language began to develop. Studies of changes in this language over time reveal that the language has moved from a structurally simpler language to a structurally more complex language (Senghas, 1995, 2000). It also appears that the differences in structural complexity show primarily in the signing of those who begin to learn the language at an early age. For example, the early form of NSL contained few verb inflections; the more recent form has such devices for marking subject-verb agreement. It is particularly those individuals who were exposed to the language at an early age that produce the verb inflections that distinguish the newer from the older form. Older learners of the newer form of the language do not master the verb inflections. Such results suggest that the changes that have occurred in NSL over time depend on young children acquiring the language. In sum, not only is language universal in the species, but the nature of the human mind ensures that wherever there are humans, there will be language. Children seem to play a unique role in creating languages with the complex grammatical systems that characterize fully developed languages. We return to this potentially special role of children when we discuss the critical period hypothesis.
The other side of the species-universality coin is species specificity. The literature on species specificity is large and messy, and there is not the space to adequately review it here. Suffice it to say that neither examination of the naturally occurring communicative systems of other species nor the several attempts to teach a language to another species have found human-like language capacities outside the human mind. The criterion on which most other systems and the most nearly successful training efforts clearly fail is syntax (Kako, 1999; Tomasello, 1994). It has also been suggested that other animals, specifically chimpanzees, lack the social interest in other members of their species that a human-like communication system requires for both invention and acquisition (Premack, 1986; Tomasello, Call, Nagell, Olguin, & Carpenter, 1994).
The Invariance and Robustness of Language Development
All normal children in anything remotely like a normal environment learn to talk. Furthermore, the course of language development is, in broad outline, constant across varying environments. These basic facts suggest to many a maturational process, the course and timing of which is determined by the unfolding of a genetic blueprint (Gilger, 1996; Gleitman, 1981). On the other hand, it could be that the universal acquisition of language is the result of universal features of human environments. A review of the literatures that describe the varied social environments in which children learn to talk suggests that all environments provide two sources of support for language acquisition: They show children that language is used to communicate with other people, and they deliver to children, through speech, data that the children use to figure out the underlying linguistic system (Lieven, 1994). Furthermore, there are differences in the rate and course of language development associated with differences in how children’s environments provide these two sources of support and in how much speech children hear. For example, in some environments children are talked to directly from birth; in others children observe and overhear conversations among others but are not engaged participants from an early age. This difference has consequences for language development. Children who are not directly talked to appear particularly precocious in the development of skills for joining the ongoing conversations of others (Bernicot & Roux, 1998; Dunn & Shatz, 1989; Hoff-Ginsberg, 1998). Children whose data is in the form of speech among others appear to begin talking by producing rote-learned chunks of speech, and they only later analyze these chunks into their structural and lexical components. Children who hear more speech addressed directly to them rely less on memorized but unanalyzed wholes (Lieven, 1994). These are effects on the style or course of language development. There are also effects on rate. Although it appears to matter more for lexical development than for grammatical development, children who hear more data develop language more rapidly than do children exposed to fewer data (Hart & Risley, 1995; Hoff & Naigles, 2002; Huttenlocher, Haight, Bryk, Seltzer, & Lyons, 1991). These demonstrated effects of the nature and amount of environmental support for language acquisition make it clear that the process of language acquisition is not solely a maturational process and, therefore, that linguistic innateness by itself is not sufficient as an explanation.
The Heritability of Language
Among children who acquire language normally, there are individual differences in the rate of language acquisition, and some children acquire language quite slowly and with difficulty. Although some of these individual differences can be attributed to differences in experience, there is also evidence that both normal individual differences and some cases of language impairment have a genetic basis. Twin studies that have looked at children past the age of 3 years using standardized tests of children’s verbal IQ, vocabulary, and reading ability find that about 50% of the variance among children on these measures can be attributed to genetics (Stromswold, 2001). Receptive and expressive language skills up to 24 months of age show lower heritability—between 1% and 38% of the variance, depending on the measure (Reznick, Corley, & Robinson, 1997). Ganger, Pinker, Chawla, and Baker (2002) used the twin study method to assess the heritability of individual differences in the timing of the achievement of two milestones of language development: the achievement of a 25-word productive vocabulary and the production of first word combinations. They found, as did earlier studies, that the heritability of vocabulary was low. In this study, 11% of the variance was attributable to genetics. In stark contrast, they found that the timing of syntactic development was highly heritable—82% of the variance was attributable to genetics. A larger scale twin study also found that the heritability of grammatical development was higher than the heritability of lexical development, although that study found lower heritability of grammatical development (39%) and higher heritability of lexical development (25%; Dale, Dionne, Eley, & Plomin, 2000).
Evidence of environmental effects on language development provides converging evidence for the greater heritability of grammatical than of lexical development. Hoff-Ginsberg (1998) found grammar to be minimally susceptible to environmental influence and vocabulary development more so. The method of studying the influence of the environment was, in this case, to investigate the effects of family socioeconomic status (SES) and birth order on language development. It is well established that children in high-SES families hear more speech than do children in middle-SES families (Hoff, Laursen, & Tardif, 2002), and it is reasonable to assume that firstborn children have more opportunity for one-to-one speech than later born children. Thus, these two variables serve as proxies for language experience. Vocabulary development was strongly affected by both birth order and family SES, with firstborn children and children from high-SES families showing larger vocabularies; grammatical development was affected only by birth order, again with firstborns showing more advanced language (Hoff-Ginsberg, 1998). It appears that grammatical development, more than lexical development, may be the result of the unfolding of a genetic blueprint; vocabulary development is more paced by environmental factors.
Evidence that impairment of the normal ability to acquire language has a genetic basis includes findings that language impairment runs in families, that monozygotic twins are more likely to be concordant for language disorders than are dizygotic twins, and that adopted children with languageimpaired biological relatives are more likely to be language impaired than are adopted children with no language impairment among their biological relatives (Eley et al., 1999; Stromswold, 1998). In one well-studied family, 16 out of 30 family members were seriously language impaired, and the inheritance patterns suggest that a single dominant gene is responsible (Gopnik & Crago, 1991).
In sum, although it is clear that differences in the environments that children experience contribute to the observed individual differences in language development, work on the genetics of language development makes a strong case for a genetic contribution as well—both within the normal range of variation and, even more so, in cases of atypical development. Syntax seems to fit the biological model of language better than other aspects of language development do. Although this work establishes that there is something genetically based that determines the pace of syntactic development, it does not clearly reveal what that something is—nor how domain specific it is. We know the LAD is a biological entity, but we still do not know how it works.
The Neurological Underpinnings of Language and Language Development
In making the argument that language is innate, Chomsky has referred to a language organ in the brain. If there is such a thing, it would seem to be located in the left cerebral hemisphere. It has been known since the nineteenth century that damage to the left side of the brain disturbs language functions, whereas damage to the right side typically does not. This is true even for deaf signers. Although the right hemisphere is primarily responsible for processing visual-spatial information, if that information is linguistic information, it is handled predominantly by the left hemisphere. A wealth of other sources of data also suggests that language is predominantly a left-hemisphere function. Patients with a split corpus callosum can label objects put in their right hand or presented to their right visual field, but they cannot label objects in their left hand or left visual field (e.g., Gazzaniga, 1983). Dichotic listening tests with normal, intact adults show a consistent right-ear advantage for speech stimuli (Springer & Deutsch, 1981). Research that uses scalp electrodes to measure eventrelated potentials or brain imaging techniques to measure cortical activity also has found greater left-hemisphere activity associated with language processing (e.g., Mazziotta & Metter, 1988). There are, however, some problems with drawing the conclusion that the language organ resides in the left cerebral hemisphere. Other parts of the brain also contribute to language processing. Patients with righthemisphere damage have difficulty understanding jokes, sarcasm, figurative language, and indirect requests (Weylman, Brownell, & Gardner, 1988). They have difficulty understanding linguistic units that have more than one meaning, and they fail to use broad contextual information in the interpretation of connected discourse (Chiarello, 1991).
The literature provides a couple of suggestions for describing the contributions of the right and left hemispheres to language. Studies of event-related potentials in intact patients show that the right hemisphere is activated by semantic processing, whereas the left is activated primarily by syntax processing (Neville, Nicol, Barss, Forster, & Garrett, 1991). Studies of the language abilities of patients who have had their left hemispheres entirely removed (because of severe pathology) similarly suggest that the right hemisphere can support many language functions but that the left hemisphere is necessary for normal syntax. Together, the findings suggest that for adults the right hemisphere is involved in semantics and pragmatics but that syntax is the province of the left hemisphere. The story becomes a little more complicated if we look at children. The right hemisphere seems to be more important for language acquisition than for language processing once language is acquired. If brain damage is suffered in infancy, prior to language acquisition, right-hemisphere damage is more detrimental to future language acquisition than is left-hemisphere damage (Stiles, Bates, Thal, Trauner, & Reilly, 1998). Another difficulty with concluding that the left hemisphere is the language organ is that although the left hemisphere is primarily responsible for language, it is not necessarily dedicated specifically to language. It has been suggested that the left hemisphere is specialized for executing well-practiced routines (see Mills, Coffey-Corina, & Neville, 1997). The finding that experienced musicians show a right-ear (i.e., left hemisphere) advantage for music stimuli, whereas naive listeners show a left-ear (i.e., right hemisphere) advantage for music stimuli (Bever & Chiarello, 1974) is consistent with this hypothesis. There is additional evidence that the relation between brain localization and language has to do with practice in the findings from children and adults that the neural representation of a bilingual’s two languages differ as a function of language proficiency (Conboy & Mills, 2001; Perani et al., 1998).
The Critical Period Hypothesis
Many examples of species-specific, biologically based learnings have the characteristic that the learning must occur during a biologically determined window if the learning is to be successful. Thus, regarding language development as a biological phenomenon suggests the hypothesis that there is a critical period for language acquisition. Seemingly, supportive evidence is obvious and widely available. When families immigrate to a new language community, it is commonly and reliably observed that the young children in the family acquire the new language well and eventually are indistinguishable from native speakers. Older children and adults acquire the language less well, and they typically never achieve native-like proficiency. Young children recover from aphasia following brain damage more rapidly and more nearly completely than do older children and adults. There is also the case of “Genie,” a child who was kept isolated and thus deprived of language until the age of 13 (Curtiss, 1977). Despite both exposure and training after her discovery, Genie never achieved normal language. Finally, there appear to be differences in how and where the brain processes language depending on the age at which the language is acquired (Kim, Relkin, Lee, & Hirsch, 1997; Weber-Fox & Neville, 1996), although differences in proficiency may be an alternative explanation. If such phenomena have a biological explanation, it should be found in some biological event. Puberty is often proposed as that event: The hypothesis is that the onset of puberty closes the window on the period during which the brain is best able to acquire language.
Not all the data fit that hypothesis. First, the advantage of being young is seen only in the level of ultimate language achievement. Older learners actually make more rapid progress than do younger learners during the first year in a new language community—provided that the opportunities are roughly equal (Snow & Hoefnagel-Hohle, 1978). Second, the advantage of younger second-language learners over older second-language learners in ultimately achieving native-like competence does not abruptly end at puberty but continues past age 20 (Birdsong, 1999). Finally, that advantage may not be so much a function of differences between younger and older brains as much as it is a function of differences between the experiences of younger and older arrivals in the new language community. Children attend school in the new language, whereas adults must do work that their limited language skills allow, thus limiting their exposure to the new language. Additionally, older children and adults have stronger first-language skills and are more likely to continue to read material in their first language than are young children. A study of Chinese immigrants to New York City found that younger children were more exposed to English than were older children during their first year in the United States, and by the end of the year Chinese was no longer preferred over and dominant relative to English for those younger children. The older children, in contrast, were exposed to English less, and at the end of their first year their preferred and dominant language was still Chinese (Jia & Aaronson, 1999). This provides an alternative explanation for why the younger children’s English proficiency was greater than that of the older children.
Although the data on second-language acquisition seem to provide less conclusive support for the critical period hypothesis on close examination than they do at first blush, other evidence remains. Recovery from aphasia is better at younger ages, and children have made unique contributions to the process of grammatical expansion in NSL. As long as these data hold, it does seem to be the case that maturation affects the way the brain acquires language. However, the strongest sort of critical period hypothesis—that the window of opportunity for language acquisition is shut by a single biological event—is not supported. This is not necessarily telling evidenceagainsttheclaimthatlanguageisbiologicallybased. First, puberty is not a single, well-defined biological event. Second, there is evidence that even in the animal world, biology alone does not determine when biologically prepared learning can occur. In songbirds, for example, access to social interaction, as opposed to mere exposure to mature song, extends the sensitive period for song learning (Nelson, 1997).
Language as an Evolved Adaptation
Consonant with the hypothesis that the capacity for language is domain specific and modular, it has been proposed that this capacity was selected for in the course of human evolution to serve its particular function. According to this view, humans have language because having language gave some of our hominid ancestors an advantage in survival and reproduction over those who did not have language (Bloom, 1998; Pinker & Bloom, 1990). This view puts language in the same category as upright posture and bipedal locomotion as evolutionary achievements writ into the human DNA. Further arguments have been made for why language would have been useful and why language evolved to have the particular structure that it does. With respect to the value of language, it has been argued that language is useful to humans because of humans’ unique social characteristics. Language is useful only to a species whose members are interested in communicating with each other, and a system as complex as human language is more useful than calls and hoots only if the interacting members of the species are interested in exchanges of information more complex than food locations and predator warnings. With respect to why language has its particular structure, it has been argued that this level of complexity is necessary for exchanges of the sort of information that is important for human survival. For example, a mechanism for embedding one clause in another is necessary to communicate the distinction between a region that has animals you can eat and a region that has animals that can eat you (Pinker, 1994). There are also arguments for how the anatomical changes that accompanied upright posture, namely the lowering of the larynx in the throat, gave rise to some of the structural features of human language (Carstairs-McCarthy, 1999).
There is no contrary argument that the human capacity for language is not part of human nature, acquired in the course of human evolution. Rather, the argument has to do with domain specificity. One contrary argument is that language evolved as a result of quantitative changes in several preexisting mental abilities (Bates, Thal, & Marchman, 1991; Lieberman, 1975). That is, the human capacity for language is a “new machine built out of old parts” (Bates et al., 1991, p. 35). Another contrary argument is that language is a byproduct of general increases in the computational power of the brain—language is just one more thing that humans do with their great intelligence (Chomsky, 1982; Gould & Lewontin, 1979).
Summary and Conclusions
If we ask to what degree language development demonstrates hallmark features of a biological process, we are led to conclude that language is an intrinsic part of human nature. Just what makes language so, however, is unclear. Language is certainly universal in the species and is species specific, but the rate and course of language acquisition are paced and shaped to a significant degree by environmental influence. Among the several components of language, grammar seems most robust—the rate of development is most heritable, and environmental effects are correspondingly weaker. Vocabulary development appears to be very influenced by access to input. Additional support for the notion that grammar is what is really biologically based about language comes from neurolinguistic research. When we look for a dedicated language organ, we find the left hemisphere carrying out the grammatical functions of language. Meaning and use appear to be less isolable functions. Most of the data that address the critical period hypothesis are inconclusive. If there is a critical period, its boundaries are fuzzy and the mechanism unclear. Yet there does seem to be an advantage to having a left hemisphere available to take on language functions, and the availability of the left hemisphere may decline with age. Furthermore, even accepting the conclusion that grammar is a biologically based human capacity that resides in the left cerebral hemisphere does not quite answer the question about the nature of that capacity. It is possible that the capacity is dedicated specifically to language. The data also fit an account in which the left hemisphere is particularly good at the kind of processing that grammar requires. Finally, the biologically based characteristics of humans on which the human capacity for language depends seem also to include the social nature of humans—although social influences are not typically included in biological models. Language is created or acquired only when there is the opportunity for interaction with other humans. This is not to say that language is entirely a social process—that issue will be taken up in later sections—but that the social nature of humans is a necessary ingredient.
Language Development as a Linguistic Process
If one starts from the premises that (a) all adults have knowledge of a complex, abstract, formal system that is what allows them to be competent speaker-hearers of a language and that (b) the information available in input and the learning mechanisms available to children are inadequate for achieving that system from an initial state of no linguistic knowledge, one is led to the inescapable conclusion that linguistic knowledge must be innate. This is basically the worldview of the generative grammar approach to language acquisition.
The task for research, then, is to describe just what is innate. This enterprise is led by linguistic analysis of the end state of language acquisition. The resultant description of linguistic knowledge, that is, Universal Grammar (UG), is then attributed wholesale to the child. Two explanatory problems remain, however. One is that children acquire different languages, depending on the language they hear, and accommodating that fact requires some mechanism that allows children to acquire the particular language to which they are exposed. The other is that young children do not demonstrate the competence that they are posited to possess, requiring some explanation for apparent developmental change.
To handle the first problem, UG allows options. It consists of a set of principles that are true for all languages and a set of parameters on which languages vary. For example, in some languages, such as English, sentence subjects must be expressed, whereas in other languages, such as Spanish, they need not be. In Spanish, one can say the equivalent of Goes to school whereas in English it would be necessary to say He goes to school. Languages that allow pro-drop, as it is called, also allow sentences that are the equivalent of Raining whereas in English one would need to say It’s raining. Because some languages allow pro-drop and others do not, this is a parameter in UG that children must set on the basis of their language experience. The argument is that children learning English must hear sentences such as It’s raining, and then use that input to set the pro-drop parameter. Less work has been done in applying a universal phonological theory to acquisition, but the structure of the problem is the same: There exists a set of universal rules (in the standard theory) or constraints (in the more recent optimality theory; see Bernhardt & Stemberger, 1998). The function of these rules or constraints is to mediate between the underlying representations of words and the surface form of words, just as syntax mediates between the underlying representation of sentences and the surface form of sentences. As it does for syntax, the linguistic approach assumes that rules or constraints cannot be learned from input. Language variation is handled by positing that a universal set of rules or constraints is provided innately, and children use input to learn which rules apply or to set the rankings of the constraints (Bernhardt & Stemberger, 1998).
This approach has encountered both theoretical and empirical problems. As linguistic theories both Principles and Parameter Theory and Optimality Theory are still in development. As a theory of acquisition, Optimality Theory has been too little tested to judge, as work in this area has only just begun (e.g., Dinnesen & O’Connor, 2001). Work applying Principles and Parameters Theory to explaining the acquisition of syntax has longer tradition, and, thus far, the proposal that children use input to set parameters has not found empirical support in data on language acquisition (Maratsos, 1998). Although future work in linguistic theory may prove more successful at describing language in terms of parametric variation and ranked constraints, Maratsos (1998) pointed out that there is a problem, in principle, with the attendant acquisition theory. The notion that a single instance of
input would serve as a trigger for parameter setting requires that input be error free, and it certainly is not. Furthermore, a parameter-setting approach in which the role of input is merely that of a trigger predicts that the amount of input should not matter a great deal to the acquisition of any parameterized aspect of language. Contrary to this prediction, there is evidence that a component of grammatical knowledge that is held to be part of UG, that-trace phenomena, is acquired earlier by children with more language experience and later by children with less (Gathercole, 2002). There may also be a solution to this problem if parameter setting is not accomplished on the basis of a single instance of input, but requires an accumulation of evidence. This, however, would be a very different sort of model than the current Principles and Parameters theory.
Turning to the problem of apparent developmental change, we find that two solutions have been proposed. One begins with the Continuity Hypothesis (Pinker, 1984), according to which all of UG is in the child’s grammar from the beginning. This requires explaining differences between children’s and adults’ grammatical performance in terms of factors other than knowledge of the grammar. Lack of lexical or pragmatic knowledge or processing limitations that interfere with performance are often invoked. The other proposed solution is the Maturation Hypothesis, according to which some elements of UG become available only later in development. Like permanent teeth and secondary sexual characteristics, these elements of grammar are provided innately, but they are not manifest until their developmental time arrives (Wexler, 1999).
This manner of reasoning and the research it generates can be illustrated with respect to one of the principles of UG. Among the principles of UG is a set of principles called binding principles, which deal with the relations between elements in a sentence. Binding Principle B pertains to how pronouns are related to nouns. Stated formally, Principle B says that personal pronouns must not be coindexed with a c-commanding noun phrase (NP) in the local domain. C-commanding within the local domain refers to a particular structural relationship between parts of a sentence. What Principle B means is that pronouns such as he and her cannot refer to a noun that is in the same part of the sentence structure as the pronoun. In the sentence Grover is patting him, him cannot mean Grover. (This is in contrast to PrincipleA, which applies to reflexives, with the consequence that in Grover is patting himself, himself must refer to Grover.)
The reason that Principle B has generated so much research is that children disobey this principle up to a very advanced age—even at 8 years—and well after they demonstrate knowledge of Principle A (Thorton & Wexler, 1999).
Language Development as a Linguistic Process
This late adherence to a basic principle of UG contradicts both the basic premise of the UG approach, that universal principles are provided innately, and the Continuity Hypothesis, that all of UG is available from the beginning. In fact, many possible explanations of children’s seeming ignorance of Principle B are problematic for the UG view. For example, one explanation is to say that Principle B is not innate after all but must be learned from experience and that the learning process takes time. The UG view rejects this explanation and in general rejects learning sorts of accounts of developmental change on the grounds that there is no way for input to give children the information they need to arrive at Principle B. That is because Principle B, like the other principles of UG, makes reference to notions like binding and c-command, and there is no way for children to get such abstract notions from input (Chien & Wexler, 1990). This is an example of how the formal description of grammatical knowledge defines the acquisition problem differently than another sort of description of grammatical knowledge might.
Another possible explanation of children’s late adherence to Principle B is to reject the Continuity Hypothesis but save UG and assert that Principle B simply matures late. The problem with this is that the whole theory requires that Principle B be on line early in order to play a necessary role in guiding other learning (Chien & Wexler, 1990). There have also been a variety of proposals regarding other factors that might interfere with children’s demonstrating knowledge of Principle B. It could be that the processing demands of the task make it difficult for children to apply their grammatical knowledge or mislead the children in some way so that children’s poor performance is a function of the task, not a lack of knowledge (Grimshaw & Rosen, 1990; Grodzinsky & Reinhart, 1993). Alternatively, it could be that children lack the pragmatic knowledge to recognize when Principle B applies.
The argument that children do know Principle B but violate it for extragrammatical reasons comes from studies in which children demonstrate that they do adhere to Principle B under some circumstances, but not others. For example, if the children interpret a sentence such as The cow is scratching her to mean the cow is scratching herself because of ignorance of Principle B, then they also should interpret Every cow is scratching her to mean many cows are all scratching themselves. Children do not do that, however.This suggests that the interpretation of the first sentence is not the result of ignorance of Principle B, but the result of some other lack of understanding. There are circumstances in which Principle B does not apply. In the sentence, She’s wearing Alissa’s clothes, Principle B requires that she not be Alissa. However, if the circumstance is that someone who looks a great deal like Alissa but who has a different haircut walks by and the identity of this person is the topic of conversation, it is possible that she does mean Alissa in the second sentence in the following sequence: She must be Alissa; she’s wearing Alissa’s clothes. One argument for why relatively old children disobey Principle B is that they have experienced exceptions such as this and they do not know that it is an exception allowed only in certain pragmatic contexts. By this account it is a lack of pragmatic understanding, not a lack of grammatical knowledge, that leads to Principle B violations.
At present there are several proposals for how best to explain the seeming late mastery of Principle B. Resolving the issue is not crucial for the present purpose because the aim of the foregoing summary of work on Principle B was not to explain why 6-year-olds sometimes interpret Grover is patting him to mean that Grover is patting himself. Rather, the purpose was to illustrate the kind of research and argument that the UG view of language acquisition generates. Although there is disagreement within the UG approach as to how to explain apparent developmental change, there is consistent agreement with the basic premise that UG must be innate.
The work on Principle B is an example of research within the UG approach that makes very little contact with other approaches to language development. Although a great deal of UGworkisofthissort,therearesomequestionsthathavebeen tackled by both UG and other approaches to the study of children’slanguage.Severalstudieshavefoundthatbeforetheage of 3 years children do not use the verbs they know in all their possible syntactic and morphological environments (Bloom, Lifter, & Hafitz, 1980; Tomasello, 1992; Valian, 1991). It is a challenge to the claims of the UG approach that children, who are asserted to have complete knowledge of the grammar, produce only a subset of the constructions the grammar allows. The UG explanation is that performance factors impede the expression of the underlying competence. Forexample, Valian (1991) (Study 5) found that early in development children produce few verbs with direct objects but that the proportion of verbs used with a direct object increases from 11/2 to 3 years. Valian (1991) interpreted the effect of age as evidence of the role of capacity limitations. Her argument was that the syntactic knowledge underlying the production of direct objects is there from the beginning, but for very young children it is too cognitively demanding for children to actually produce such sentences. As processing capacity increases with age, such performance constraints lessen. We consider an alternative account of these data in the next section.
Innate Lexical Constraints
There is also a learnability approach to lexical development. It begins, like the argument for innate grammar, with an assertion of the impossibility of learning. This assertion was famously made by the philosopher Quine (1960) and is known as the Gavagai problem. Imagine that you are a linguist studying a newly discovered tribe. A rabbit runs across the field, and a tribesman shouts Gavagai. How do you decide what Gavagai means? It could mean “running thing,” “whiteness,” “furriness,” “dinner if we’re lucky,” “animal,” “mammal,” or even “rabbit.” The point is that in principle, the possible meanings are infinite. Despite this problem, children are good at learning words. By the time children are 6 years old they have vocabularies of up to 14,000 words. Assuming that word learning begins at about 12 months, that works out to a word learning pace of nearly eight words per day. Furthermore, experimental work has demonstrated that very few exposures to a novel term are sufficient for children to form at least a partial entry in their mental lexicons (Dolloghan, 1985; Woodward, Markman, & Fitzsimmons, 1994). Thus, somehow, the infinite possible meanings of a new word in context that bothered Quine do not bother young, language-learning children.
To solve the problem of how children arrive at meanings readily, when the environment does not clearly indicate those meanings, it has been proposed that children come to the word learning task with innate constraints on the kinds of hypotheses they consider. A great deal of research has been aimed at discovering just what those constraints are. One constraint or principle that appears to guide children’s inferences about the meanings of newly encountered words is the whole-object principle. This leads children to assume that words refer to whole objects rather than to a part or property of an object. This eliminates “whiteness” and “furriness” as possible meanings of Gavagai. The existence of such an assumption is supported both by evidence from word-learning experiments (Markman & Wachtel, 1988; Taylor & Gelman, 1988; Waxman & Markow, 1995) and by errors that young children make. For example, it is not uncommon for very young children to think that hot is the label for stove, given the common experience of hearing, Don’t touch it; it’s hot in reference to the stove.
Another proposed word-learning principle is the taxonomic principle, according to which words refer to things that are of the same kind. This assumption—it is proposed— helps the child figure out what else, other than the particular whole object being labeled, is included in the meaning of the new word. When the child hears the word dog in the presence of a dog (and assumes that the whole dog is being referred to because of the whole-object principle), the taxonomic principle leads the child to think that dog will also refer to other dogs, but not to things that are thematically related to dogs, such as collars, leashes, or bones. Again, the evidence for this assumption comes from both the fact that in natural language use children tend correctly to extend the meanings of the words they learn and experimental demonstrations that despite a tendency to form thematically related categories, when children are presented with a new word, sud with a picture (e.g., a picture of a dog) and asked to find another one that is the same as this sud, children pick another picture of a dog. When simply asked to find another one that is the same—without a new word being introduced—preschool children are more inclined to pick the picture of dog food (Markman & Hutchinson, 1984). Another principle, the mutual exclusivity principle, is the principle that different words refer to different kinds of things. So, for example, members of the category labeled dog do not overlap with members of the category labeled cow. In word-learning experiments, when presented a novel word and an array of objects that includes several objects for which they have labels and one for which they do not, children will take the new word to mean the object for which they have no label. That is, they appear to assume that the new word cannot be a synonym for any of the words that they already know (Markman & Wachtel, 1988; Mervis & Bertrand, 1994).
These principles work in concert, so that the mutual exclusivity assumption provides a basis for overriding the whole object principle, which children must do in order to learn terms for parts and properties of objects. A child who knows the word rabbit and hears the word furry will not take furry to be a synonym for rabbit but will look for something else to be the referent of the new term. The alternatives to the proposal that word learning is constrained by innate, linguistic principles are (a) that these principles operate in word learning but are themselves learned (Nelson, 1988) and (b) that the assumptions that guide word learning are not specifically linguistic but have a social-pragmatic basis (Clark, 1997). It has also been argued that the process of word learning involves more than mapping sounds onto referents and that multiple learning procedures are brought to bear on the task of building a lexicon (Hoff & Naigles, 2002).
Summary and Conclusions
It is difficult to evaluate the contribution to the general understanding of language development made by the UG approach to grammatical development. The reason for this difficulty is that this work is so isolated from the rest of the field that one must either buy the UG account or reject it as a whole package. This isolation is not a scientific necessity, but rather a sociological fact. There is very little discussion between linguists and nonlinguists because researchers from the two disciplines do not ask the same questions or even use the same vocabulary. There are, of course, exceptions. The work reviewed on the issue of whether children have the category V is an exception, and there is an interesting and empirically informed debate on this topic. Similarly, in the domain of lexical development, the source of and necessity for constraints on the hypotheses children generate has motivated a great deal of research, and that research has advanced understanding of the process of word learning. For example, mixed accounts of lexical development incorporate some innate principles along with other bases for word learning (Golinkoff, Mervis, & Hirsh-Pasek, 1994; Hollich, Hirsh-Pasek, & Golinkoff, 2000).
Even without such integration, however, the linguistic approach makes a contribution to the overall enterprise of understanding language development by forcefully making two points: (a) Grammar is complex, and it is not so easy to show how some of the more abstract aspects of grammar could be derived from input; and (b) children may have more grammatical knowledge than they demonstrate in their speech, because producing speech requires adjunct competencies such as memory and pragmatic understandings, which may be limited in children.
Language Development as a Social Process
The approach that considers language as a social phenomenon and language development as a social process begins with very different assumptions than does the previously discussed generative grammar approach. Whereas the generative grammar position asserts that the output of the LAD is a complex, abstract system whose properties are unrelated to the communicative function of language and are not apparent in the surface form of sentences, the social approach, also termed the social-pragmatic view, asserts that language is much simpler (Tomasello, 1992, 2000). The social-pragmatic approach relies on Cognitive-Functional linguistic theory, according to which grammatical devices are not hidden things like c-command and binding but are things like word order and case marking endings on words. Not only are these grammatical devices directly observable, but their functions are related directly to meaning. That is, word order and case markings indicate notions such as agent of action and object of action that are part of everybody’s cognitive understandings of events.
Second, in the social-pragmatic view, the input is richer than in the generative grammar view. According to the socialpragmatic view, it is crucially important to the process of language development that language is learned in the context of interaction with others, in routines such as feeding and dressing and, in some cultures, interactive games, book reading, and more. The significance of these repeated routines is that they create a shared referential context within which the language of the adult makes sense to the prelinguistic child (Carpenter, Nagell, & Tomasello, 1998). Third, the child’s learning mechanisms include the all-important capacity for what Tomasello has termed cultural learning (Tomasello, 2000; Tomasello, Kruger, & Ratner, 1993). Cultural learning involves imitating others, but it is not uncomprehending mimicry. Rather, cultural learning consists of learning to reproduce the behavior of others for the purpose of achieving the same goal or performing the same function that the learner understands to have been the goal or intended function when that behavior was produced by another. Cultural learning is possible only after around 9 months of age, when children develop the capacity for secondary intersubjectivity, which allows them to coordinate their attention with both the person who is talking to them and the objects or events that are being talked about. This is prerequisite to cultural learning because only with this capacity can children share another’s perspective on the world, thereby gaining nonlinguistic access to the communicative intentions of others. Learning language, on this view, is learning to express one’s communicative intentions with sequences of sounds that one has previously heard uttered by others to express those same communicative intentions.
In sum, compared to the generative linguistics approach, the social-pragmatic approach assumes that the infant who acquires language is smarter and that the language system that is acquired is simpler. Viewed this way, language is learnable without recourse to stipulation of innate knowledge. In the next sections we consider what this view has to say more specifically with respect to the acquisition of grammar and the lexicon, and we consider some of the arguments that have been made against those proposals.
A Social-Pragmatic Account of Grammatical Development
The gist of the social pragmatic approach to explaining grammatical development is to redescribe both the nature of the child’s syntactic competence and the adult syntactic competence eventually achieved so that they are states of knowledge that can be achieved using information available in social interaction and domain-general mechanisms of learning. (There is no social-pragmatic account of phonological development.) The redescription of children’s competence is done through analysis of children’s spontaneous speech and performance in experimental tasks. The redescription of adult competence is made by referring to cognitive-functional linguistics.
The crucial point in the redescription of child competence is the claim that children do not have a fully productive system. It is that productivity, or generativity, that motivates the abstract system that is generative grammar. Children, it is argued, have only a repertoire of constructions that they have memorized as a result of hearing them in input. Central to this argument is the claim that children do not have the grammatical category V, because sentences are built around verbs. Without the category children cannot have fully productive rules for sentence generation. Instead, children have verbspecific frames that they use to construct sentences. Children do have a noun-like category, and thus they achieve some limited productivity in their speech by substituting nouns in their memorized, verb-based frames.
Recall from the previous section, the observation that children initially do not use the verbs that they know in the full range of environments that the grammar allows. Rather than explaining this as a reflection of performance limitations, as the UG approach does, the social-pragmatic approach argues that children have limited knowledge about how individual verbs can be used. In essence, children know only what they have heard. In support of this view, there is evidence that children use the same frames with each verb as their mothers use (Theakston, Lieven, Pine, & Rowland, 2001). Additionally, there are findings from several experiments in which children are taught new verbs and then asked questions designed to elicit use of those verbs in different sentence structures. The consistently obtained result is that children under 3 years tend not to use their new verbs in utterances that go beyond the way they have heard those verbs used by adults. This contrasts with the finding for novel nouns, which children do use in new combinatorial structures (Tomasello, 2000).
Eventually, of course, children do use verbs productively, and the social-pragmatic theory must account for that. Part of the social-pragmatic solution is the proposal that children move to a system of more general, productive constructions from an item-based set of constructions by noticing patterns among the item-based constructions and combining them. Consistent with this sort of mechanism is that some measures of productivity increase gradually from 2 years to 8 years (Tomasello, 2000). The social-pragmatic account never gives itself the task of explaining how children achieve the autonomous, abstract grammar of the generative linguistic approach. This is where the redescription of adult competence comes into play. Adult linguistic competence, in this view, is also just an inventory of constructions—the difference is that some of them are quite general and productive. The adult grammar can end up with some of the structural apparatus of standard grammars, provided that those structures relate to specific communicative functions. So, for example, syntactic constituents such as noun phrases can be arrived at from input because noun phrases refer to identifiable referents. To paraphrase an example from Tomasello (2000), your papers is a coherent constituent because it serves a single referential function. The importance to acquisition of children’s understanding of speakers’ communicative intentions is made clear by this analysis. All the units of language structure correspond to communicative functions. Thus, understanding the communicative functions of utterances leads inevitably to understanding structure. In this view, no other route to grammar is necessary.
The counterarguments to the social-pragmatic view of the acquisition of grammar are essentially three: (a) Whereas the social understandings that are argued to be the basis of language development are important and perhaps even necessary for some aspects of development, they are not sufficient to account for acquisition; (b) the social-pragmatic position and Cognitive-Functional linguistics underestimate the complexity and functionally independent nature of morphosyntax (Maratsos, 1998; Shatz, 1992); and (c) the data do not quite support the straightforward input-dependent, limitedproductivity account offered by the social-pragmatic view. To elaborate the last point, diary data on the first uses of 34 common verbs by eight different children show that many verbs do in fact appear in multiple structures within children’s first 10 uses of those verbs, suggesting some early productivity (Vear, Naigles, Hoff, & Ramos, 2001). In addition, although children do tend to use main verbs in the particular syntactic structures in which they have heard those verbs used (deVilliers, 1985; Theakston et al., 2001), this is not the case for auxiliary verbs. One of the most robust input-acquisition relations in the literature is that the frequency in input of questions that prepose auxiliaries (e.g., Can you eat your breakfast now?) is related to children’s acquisition of auxiliaries, but when the children use auxiliaries it is first in declarative forms (e.g., I can do it; Newport, Gleitman, & Gleitman, 1977; Shatz, Hoff-Ginsberg, & MacIver, 1989). This particular datum and the complexity of what children acquire more generally argue that internal mental processes operate over input and transform it into syntactic knowledge that is used in creating novel productions. The socialpragmatic approach to the acquisition of grammar is mute with respect to what those internal mental processes might be.
Social-Pragmatic Approach to Lexical Development
The gist of the social-pragmatic explanation of lexical development is the assertion that children can figure out what newly encountered words mean because they know what the other speakers’communicative intentions are. Early formulations of the social-pragmatic proposal argued that the recurrent social interactions between mother and child establish each participant’s intentions through a given routinized activity, allowing the child to predict “where the adult’s attention is currently focused and where it is likely to be focused next. Therefore, any language the adult may use in such a context is likely to be immediately meaningful to the child” (Tomasello & Todd, 1983, p. 199). Additionally, if in nonroutinized interaction, mothers talk about the aspects of the activity that the child is focused on, then the meanings the child is harboring should be consistently expressed. Thus, the social-pragmatic argument is that by virtue of either routinization or maternal attentiveness children know what their mothers are saying without understanding the language, and they can use that nonlinguistically acquired knowledge to figure out the meaning of the language they hear.
More recently, the social-pragmatic proposal has focused on the social cognitive abilities and inclinations of children (Akhtar & Tomasello, 2000; Baldwin, 2000). Children are not atthemercyofadults,followingtheirattentionalfocusinorder for word meaning to be made transparent. Rather, children havetheabilitytodiscerntheirmothers’communicativeintentions.According to this view, word learning begins once children understand others as intentional agents, assume some communicative intention behind the vocalizations others make, and successfully figure out what those communicative intentions are.Although the earlier and later formulations differ in whether the mother or the child contributes the requisite social-pragmatic skill, in both formulations mutual engagement or joint attention is prerequisite to word learning.
The sort of empirical findings cited in support of this view include findings that maternal responsiveness is positively associated with child vocabulary development (Tamis-LeMonda, Bornstein, Kahana-Kalman, Baumwell, & Cyphers, 1998), that mothers who follow their child’s lead have children with larger vocabularies (Akhtar, Dunham, & Dunham, 1991; Harris, Jones, Brookes, & Grant, 1986; Tomasello & Farrar, 1986), and that the proportion of time mothers and children spend in joint engagement predicts vocabulary growth (Carpenter et al., 1998; Tomasello & Todd, 1983). Another important piece of this argument is evidence that children have the ability to use nonverbal cues such as eye gaze to determine a speaker’s focus of attention and that they infer that speakers are talking about their own focus of attention, even when it differs from the child’s (Baldwin, 1993).
The foregoing arguments address how children can correctly map a newly encountered word onto its intended referent. There is also a pragmatically based argument for how children learn multiple terms for the same referent (Clark, 1997). That is, the same creature can be the dog, our pet, or Rover, depending on the purpose for which one is labeling him. Children are disinclined to take these different labels to be exact synonyms because they have the principle of contrast, that is, the pragmatic understanding that if a speaker chooses a different word, it indicates a different meaning. The innate constraints approach explains the same behavior with reference to the mutual exclusivity principle. There is evidence consistent with both views. Absent any pragmatic information, children are reluctant to accept new labels for an already-labeled referent (Merriman & Bowman, 1989); however, very young children can learn multiple labels for the same referent if they understand the multiple communicative purposes or perspectives that lead to different word choices (Clark, 1997). Adults also provide help to the would-be word learner through pragmatic directions. To illustrate, Clark offers a hypothetical example in which some birds are flying overhead and the adult speaker first says look at the birds and then points to one bird followed by another, saying This one is a sparrow; this one here is a crow. By first not individuating and later individuating the referents, the speaker indicates to the child the perspective shift that accompanies the shift in the hierarchical level of the label provided. Although the foregoing example is hypothetical, there are data to the effect that mothers indicate when they are talking about a group labeled by a superordinate category and that simple labeling tends to be used with basic-level or subordinate categories (Callanan, 1985). There are also spontaneous speech data from very young children attesting to their use of multiple labels for the same item (e.g., food and cereal, animal and tiger; Clark, 1997).
The quarrel some have with the social-pragmatic approach to lexical development is similar to the exception taken with respect to social-pragmatic accounts of grammatical development. The problem is one of sufficiency. Although children may well use their social skills to infer speaker intent in episodes of mutual engagement, analysis of published studies of mother-child interaction suggests that only about 20% of the time mothers and children spend in conversation is characterized by the mutual understandings upon which the social-pragmatic view of acquisition depends (Hoff & Naigles, 2002). Furthermore, there is more to building a lexicon than mapping newly encountered words onto referents. Before that initial mapping words must be identified as such in the speech stream, and after the initial mapping there is still more for children to figure out to complete the lexical entry. The sort of information that would contribute to those other parts of the word-learning task and the sort of learning abilities that would use that information are ignored in the social-pragmatic account.
Acquiring the Social Uses of Language
Thus far we have been discussing socially based accounts of the acquisition of what are considered the core aspects of language: grammar and the lexicon. The extent to which social processes can explain those developments is very much at issue. In contrast, there are other aspects of language development for which a socially based explanation is the most obvious choice. Language development includes language socialization—learning to use language as do adult members of one’s social group. This refers to everything from learning to be polite to learning to tell a coherent story. These learnings must depend on the social context in which language is acquired because what constitutes both politeness and a coherent story vary depending on culture (Clancy, 1986; Minami & McCabe, 1995). More direct evidence for the role of social context in these aspects of language development are findings that within a culture, differences in the environmental support provided by other speakers are related to the rate of development of the ability to produce narratives (Fivush, 1991; McCabe & Peterson, 1991; Reese & Fivush, 1993).
Socially based accounts of development have little competition with respect to explaining the acquisition of communicative competence. While there are requisite and perhaps biologically based cognitive developments, it is clear that acquiring communicative competence depends on social experience. A large body of work describes children’s development of communicative competence and seeks explanation in terms of social development, cognitive development, and experience in language interaction (see, e.g., Ninio & Snow, 1999). From a distance, however, this sort of research seems not to be the main-stage event in the search for an explanation of how children learn to talk. The main stage is occupied by the study of grammar and the lexicon.
Summary and Conclusions
There is no quarrel from any quarter that language is used to communicate or that it is acquired in the context of communicative interaction. Thus, there is most certainly a social basis to language development (Shatz, 1992). At issue is the explanatory power of this social basis. The claim made by advocates of the social-pragmatic view—that learning language is merely a matter of learning to produce the sounds to express one’s intentions that one has heard others produce to express those intentions—is not particularly revolutionary.
If intentions are meanings, and it seems that they must be, then this statement is simply a paraphrase of a statement in every linguistics textbook: Language is a system for relating sound to meaning. The question is, What is the nature of that system and what does it take to learn it? According to the social-pragmatic approach, the system is fairly simple and based in the communicative functions of language. Thus, learning the system is a natural consequence of children’s social-cognitive abilities to understand others’ intentions, their desire to communicate their own intentions, and their ability to imitate the means of communicating intentions that they see modeled by others. To the extent that language is complex and has structural properties not grounded in the communicative functions of language, it will require more to be learned than social understandings and a capacity to imitate the goal-directed behavior of others.
Language Development as Domain-General Learning
It is possible, in principle, that substantial aspects of linguistic knowledge are neither innate nor achieved in the manner described by the social-pragmatic approach. Linguistic knowledge could be learned via the application of asocial and domain-general learning procedures to language input. The work in this vein does not come from a single theoretical orientation but rather comprises a variety of approaches that have in common a rejection of the assumptions that language is too complex, the input too impoverished, and the child’s learning mechanisms too weak to account for language acquisition without stipulating innate linguistic knowledge. Some, but not all, domain-general proposals also entail a description of the adult linguistic knowledge that differs from the generative linguistics account. The domain-general learning approach differs also from the social-pragmatic approach in locating the information used by language learners in the speech stream itself and in the nonlinguistic context accompanying speech rather than in socially achieved understandings. Information about utterance or word meaning that children derive through their understandings of other speakers’ intentions may be useful, but it has no special status that is different from other sources of information. Furthermore, the mechanisms for learning are more than the mechanisms of cultural learning referred to by the social-pragmatic approach. The argument for a domain-general approach to language development is made primarily with three sources of evidence: (a) studies of infants showing that language learners do indeed have powerful learning mechanisms available to them, (b) computer implementations of connectionist models of language acquisition demonstrating the sufficiency of some aspects of input for at least some aspects of language development, and (c) developmental data suggesting that children do apply general learning procedures to input in the process of acquiring language.
The Nature of Infants’ Learning Mechanisms
Recent attention to infant learning mechanisms began with a study demonstrating that 8-month-old babies can learn the statistical regularities in a stream of sounds presented for only 2 min (Saffran, Aslin, & Newport, 1996). The babies listened to a tape-recording that presented four different threesyllable “words” (e.g., tupiro, golabu) combined in random order in an uninterrupted stream. After 2 min of exposure, the babies were tested with stimuli that either recombined the same words in a different order or recombined the same syllables in a different order, violating the integrity of the former “words.” Eight-month-old babies could tell the difference. They appeared to have learned the distributional regularities in the first sequence, that is ro always follows pi which always follow tu, whereas tu can follow any of three different word final syllables. Although this finding demonstrated that babies were more powerful learners than had previously been thought, there remains substantial disagreement over how much of the burden of explaining language acquisition even this kind of learning can carry. Some have hailed this finding as evidence that babies can learn language (Bates & Elman, 1996), whereas others, including the original authors, have been quick to point out that there is more to learning a language than identifying words in the steam of speech (Pinker, 1996).
There is other evidence that babies are capable of even more sophisticated learning. That is, babies seem able not only to learn the distributional regularities among the particular sounds they hear but also to abstract a pattern that can be applied to other dissimilar sounds (Marcus, Vijayan, Bandi Rao, & Vishton, 1999). Marcus et al. (1999) presented 7-month-olds with 2-min sequences that followed an ABA pattern (e.g., ga ti ga, li na li) or anABB pattern (e.g., ga ti ti, li na na). Subsequently, the babies were able to distinguish between the pattern that they had heard and the other pattern, even when those patterns were presented using entirely different syllables. According to Marcus et al. (1999), these babies learned algebraic rules, not just statistical regularities, but there are dissenters from this view (McClelland & Plaut, 1999). The assertion that infants can learn rules has broad implications. Rules capture patterns among abstract variables, and variables can refer to any stimuli, old or new, regardless of their similarity to the stimuli that produced the learning.
Thus, a rule is highly generalizable knowledge. In contrast, statistical learning consists of memorizing the co-occurrence patterns among stimuli actually experienced. These patterns can be generalized, but there must be some physical similarity between the stimuli that produced the learning and the new stimuli to which the learning is applied. The scope of generalization for statistical learning is much narrower than for rule learning, and rule learning is therefore a much more powerful learning mechanism (Marcus, 2001).
The significance of the evidence that babies can learn rules is tied up with one of the central topics of dispute between some domain-general approaches to language acquisition and the generative linguistic approach: Is knowing language knowing a system of rules?According to traditional accounts, it is. The rules are rules for word formation (in phonology and morphology) and sentence formation (in syntax) that operate oversymbols.Thesymbolsarevariablesthatstandforabstract categories such as Noun and Verb; the rules apply to anything that is an instance of those categories. This is what allows language to be productive. According to some connectionist accounts, linguistic knowledge is not knowledge of a rule system that operates on variables. Rather, language is a function of the strength of connections among less abstract units— in some models the units are actually sound sequences.
Connectionism as a Domain-General Challenge to Nativism
Connectionism is the main domain-general challenge to linguistic nativism, and there have been computer implementations of connectionist models of aspects of the development of phonology (e.g., Plaut & Kello, 1999), of morphology (e.g., Daugherty & Seidenberg, 1994), of syntax (e.g., Elman, 1993), and of lexical development (e.g., Siskind, 1996; see Brent, 1997; Plunkett, 1998, for comprehensive treatments). The basic learning mechanism is that of establishing connections among units and adjusting the strength of those connections. The strength of the connection between two units increases each time the stimuli that activate those two units are presented together. Thus, like old-fashioned associationism, connectionist learning is a function of the contiguity of stimuli in experience. Unlike older models, however, connectionist models contain a complex internal structure that mediates between experience and learning with the result, it is claimed, that what is learned is greater than the sum of the learner’s experiences. The strongest claim of connectionism is that it offers an alternative to the tradeoff between finding the structure of language in the input or building the structure into the acquisition mechanism. Instead, structure emerges from the effect of input on the connectionist network (Plunkett, 1998; see also MacWhinney, 1999, for a more complete treatment of the notion of emergence). Another attraction of connectionist models is that they seem closer to biology than symbolic models do because we know that the brain is a set of interconnected neurons. If cognitive processing could be modeled in a system that is closer to the “wetware” of the brain, the hope is that the problem of determining how the brain represents symbols and rules could be eliminated.
Although connectionism has generated a great deal of excitement as a potential new way to explain both how language is learned and how the brain accomplishes this feat, there are also naysayers. The success of many connectionist implementations is disputed, as is their relevance to what children actually do. Also, connectionism is not an alternative to a system of symbols and rules if it is merely an implementation of symbols and rules. According to Marcus (1998), many connectionist models—particularly those that successfully mimic some aspect of language acquisition—actually contain within them nodes that stand for variables. If the connectionist model merely implements a symbolic processor, then the problem of explaining symbolic processing and the attendant problems of explaining the acquisition of symbols and rules are not solved. Furthermore, the analogy between connectionist models and the brain has been criticized as illusory. Not only are nodes not neurons, but also no one knows how the neurons in the human brain represent what humans know. For that reason, connectionism does not bring us closer to knowing how the brain represents or acquires linguistic knowledge (Fodor, 1997).
Even if connectionism does not provide an account of how children learn language using domain-general learning procedures, there are other arguments that language development depends on some sort of general learning—that neither innate linguistic knowledge nor solely social processes can be the entire story. We turn now to more specific proposals within the domains of phonological, morphosyntactic, and lexical development and to arguments that have been made on the basis of developmental data.
A Problem-Solving Model of Phonological Development
In contrast to the view of linguistic nativism—that all the universally possible rules or constraints of phonology are provided innately and selected or ranked by input, there is a widely held view that phonological development is the result of the child’s problem-solving activity (Ferguson & Farwell, 1975; Macken & Ferguson, 1983). The problem that requires solution by the child is how to match the target language given the child’s articulatory constraints. According to the problem-solving model, children figure out how to use their limited articulatory abilities to approximate the target language using their general capacities for perception, production, and problem solving. Initially children do this on a word-by-word basis, with the attested result that sounds may be produced in different ways across words. Later children arrive at a system for mapping individual sounds in the target language onto a set of sounds that they can produce. This results in consistency, but occasionally in regression for sounds that were produced accurately in only some words. Because different children hit upon different solutions to the problem, there will be individual differences among children in the phonological systems achieved. Applications of nativist models in contrast tend to focus on universals of phonological development rather than individual differences, and part of the argument among different approaches concerns how much individual variation exists.
Domain-General Processes in Morphosyntactic Development
The argument that the development of morphology and syntax are the result of general cognitive processes has been made on the basis of analysis of the task itself (Maratsos, 1998) and on the basis of evidence that children make use of information in input in acquiring syntax (Hoff-Ginsberg, 1986; Shatz et al., 1989). The task analysis consists of describing the complexity of morphological systems that are specific to particular languages and thus could not be provided innately. (In fact, the UG position makes no claim in this regard.) The question is how could children acquiring Turkish, to pick an example, learn that a particular suffix, -u, indicates patienthood of the noun to which it is attached (and, according to Maratsos, 1998, this is a simplified example of the problem of learning Turkish morphology). Children will hear sentences in which a noun has -u as a suffix, but they need to figure out from the universe of possibilities what that suffix indicates. Maratsos argued that there must be some innate constraints on the possibilities considered, but languages vary enough in what meanings get grammaticized that even an innately constrained list of possibilities would be quite long. The child would need many examples of the -u suffix in order to figure out what meaning reliably co-occurs with -u. This process of figuring out is a process of sifting through data, and it requires a great deal of data. Thus, although some innate constraints on the problem may be involved, the process is not like the learning process in UG models in which input functions as a trigger. The process is also unlike that proposed by the social-pragmatic view because neither the hypothesizing of candidate meanings nor the sifting through data has anything to do with social processes. Other speakers do not point out which of the many meanings that co-occur with a particular instance of -u is being encoded by that suffix, and the process of grinding through the data to find the right one is entirely internal to the child. Learning depends on asocial input and internal data-analytic processes operating over that input.
Add to Maratsos’s argument for data sifting the facts that languages are not perfectly regular and that speech is not completely error free. This leads to the inevitable conclusion that the internal data-sifting mechanism must be able to detect probabilistic patterns in the environment. In fact, Newport and Aslin (2000) documented children’s abilities to detect the regularities in less-than-perfect input, and they argued that this importantly enables language acquisition. The ability to exploit probabilistic information is useful to most species in many domains because, as Kelly and Martin (1994) put it, “the structure of the environment itself is often probabilistic” (p. 105). Thus, we are led to the conclusion that the acquisition of morphology depends on massive language data that feed data-analytic abilities that are domain-general.
The notions that input provides information that requires an internal data-sifting mechanism, that children use that information in learning language, and that the value of that input is independent of its communicative function are also supported by evidence from the study of the relation between properties of mothers’ child-directed speech and their children’s syntactic development. The frequency with which mothers ask questions with preposed auxiliaries, produce partial expansions of child speech, and produce partial selfrepetitions and expansions have all been found to predict the rate at which children develop some aspects of grammar (N. D. Baker & Nelson, 1984; Hoff-Ginsberg, 1985, 1986; Newport et al., 1977). Hoff-Ginsberg (1985) argued that these features of input are useful because they make the phrase structure of language salient to a distributional learning mechanism. Finally, the properties of input that provide this sort of information are unrelated to the functions to which the speech is put. Although the use of question forms is of course related to the expression of questioning as a communicative intent, it is not clear what the functional basis would be of a particular benefit of questions that involve preposed auxiliaries. Mothers’ use of partial self-repetitions and expansions has been found to be unrelated to the functions of maternal speech (Hoff-Ginsberg, 1986, 1990, 1999).
Domain-General Processes in Lexical Development
The proposal that domain-general learning processes contribute to lexical development also begins with an analysis of the task itself. Current research and theory suggest that the process of word learning consists of at least the following three, ordered, components: (a) word segmentation, (b) an initial fast mapping of the new word onto a referent, and (c) a longer, extended process of completing the lexical entry. Both the innate constraints view and the social-pragmatic view address only the process of mapping words onto referents. They have nothing to say about word segmentation or the internal data-sifting processes that must be involved in completing a lexical entry. In contrast, there is substantial evidence that the amount and nature of input contribute to the child’s lexical development. With respect to word segmentation, there is evidence that stress patterns, prosody, and repetition of words in combination with a variety of different words all contribute (see Bibliography: in Aslin, Saffran, & Newport, 1999; Morgan & Demuth, 1996). With respect to the word-to-world mapping that the innate constraints and social-pragmatic views address, there is evidence for the potential usefulness and actual use of other information in the content and structure of the utterance in which a new word appears. Computer simulations (Siskind, 1996) have demonstrated that the use of partial linguistic knowledge to constrain hypotheses, combined with the ability to extract commonalities across different situations of use, can result in lexical acquisition by a system that has no access to speaker intentions. To illustrate, knowledge of what the word ball means, combined with knowledge about what kinds of entities do what kinds of things, indicates to the learning device that if the word ball is in the utterance, then an unknown word in that utterance is more likely to mean “roll” than “eat.” Evidence that humans can similarly make inferences about word meaning from information in the utterance containing a novel word comes from Gillette, Gleitman, Gleitman, and Lederer’s (1999) simulation of word learning with human (adult) participants. The learners in this case were shown a series of silent video clips of real mother-child interactions during which a specific verb had been spoken by the mother. The participants were provided with various clues to the identity of the verb, including (a) just the video clips, (b) the video clips plus the nouns in the mother’s utterance, (c) just the nouns in the utterance, (d) just the sentence frames in which the verb was placed, (e) the sentence frames plus the nouns, or (f) the video clips, the sentence frames, and the nouns. With only the video clip information, the participants made correct identifications of the target verbs only 7.7% of the time. Each additional bit of information raised this level of accuracy significantly, until those with complete information (i.e., Condition f) reached 90.4% correct.
These foregoing studies involve simulations by computer or adult learner. There is also substantial evidence from studies of children’s performance in experiments and from studies of input correlates of lexical development that young children find sources of information in the speech they hear and in the nonlinguistic context of that speech. Those findings suggest that neither the innate constraints view nor the socialpragmatic view provides a complete account of the process of word learning. Experimental studies of young children show that they can make use of structural information in figuring out what a newly encountered word means (Goodman, McDonough, & Brown, 1998; Prasada & Choy, 1998; Waxman, 1999; see Woodward & Markman, 1998, for a summary). For example, Naigles (1990) found that given a scene in which multiple interpretations of a novel verb are possible, 2-year-olds make systematically different conjectures depending on whether they hear The duck is gorping the rabbit or The duck and the rabbit are gorping (see also Naigles, 1996, 1998). Children’s ability to use cross-situational information to converge on meaning has also been demonstrated (Akhtar & Montague, 1999), again suggesting an internal process of data sifting.
Evidence that children actually use syntax as a source of information in lexical development comes from studies of correlations between properties of input and measures of lexical development that find, for example, that the diversity of syntactic frames in which a verb appears predicts its order of acquisition (Naigles & Hoff-Ginsberg, 1998) and, more generally, that the syntactic complexity of maternal speech (indexed by mean length of utterance) is positively related to children’s vocabulary development (Bornstein, Haynes, & Painter, 1998; Hoff & Naigles, 2002). Evidence that the process of lexical development requires a great deal of data to sift through comes from multiple findings that the amount of input addressed to children is a positive predictor of their vocabulary development. Huttenlocher et al. (1991) found that the amount of speech mothers produced was a significant, positive predictor of their children’s rates of vocabulary growth over the course of 10 to 12 months. Studying the vocabulary of bilingual children, Pearson, Fernandez, Lewedeg, and Oller (1997) found that the relative sizes of 1- to 2-yearold children’s vocabularies in each of the languages they were acquiring was related to the relative amount of input in each language.
Summary and Conclusions
The evidence reviewed in this section argues that Chomsky may have overstated the case when he said that input was too impoverished and children’s learning mechanisms too weak for language to be learnable. The evidence suggests that children’s language development is to a nontrivial degree the result of children’s general capacities for learning the regularities in their environment applied to information in language input. As was the case for the social-pragmatic approach, the problem of sufficiency remains. None of the demonstrations that input is useful and used in the process of language acquisition constitutes proof that linguistic nativism is unnecessary.
In this research paper we reviewed the theoretical and empirical arguments that address the nature of the capacity that underlies language development. We organized those arguments under four headings that refer to four different premises regarding the nature of that capacity: the biological, the linguistic, the social, and the domain-general cognitive approaches to the study of language development. No approach seems sufficient alone, and each approach makes some contribution to the current understanding of how children learn to talk. The implication is that nature has equipped children with innate constraints, social inclinations and abilities, and asocial computational processes, all of which contribute to children’s language development. If this is true, then an important goal for researchers in the field of language development must be to specify not only the nature of each source’s contribution but also how these factors interact in the course of language development.
- Akhtar, N., Dunham, F., & Dunham, P. J. (1991). Directive interactions and early vocabulary development: the role of joint attentional focus. Journal of Child Language, 18, 41–50.
- Akhtar, N., & Montague, L. (1999). Early lexical acquisition: The role of cross-situational learning. First Language, 19, 347–358.
- Akhtar, N., & Tomasello, M. (2000). The social nature of words and word learning. In R. M. Gollinkoff, K. Hirsh-Pasek, L. Bloom, B. Smith, A. L. Woodward, N. Akhtar, M. Tomasello, & G. Hollich (Eds.), Becoming a word learner: A debate on lexical acquisition (pp. 115–135). Oxford, UK: Oxford University Press.
- Aslin, R. N., Saffran, J. R., & Newport, E. L. (1999). Statistical learning in linguistic and nonlinguistic domains. In B. MacWhinney (Ed.), The emergence of language (pp. 359–380). Mahwah, NJ: Erlbaum.
- Baker, C. L., & McCarthy, J. J. (Eds.). (1981). The logical problem of language acquisition. Cambridge, MA: MIT Press.
- Baker, N. D., & Nelson, K. E. (1984). Recasting and related conversational techniques for triggering syntactic advances by young children. First Language, 5, 3–21.
- Baldwin, D. (1993). Infants’ ability to consult the speaker for clues to word reference. Journal of Child Language, 20, 395–419.
- Baldwin, D. (2000). Interpersonal understanding fuels knowledge acquisition. Current Directions in Psychological Science, 9, 40– 45.
- Bates, E., & Elman, J. (1996). Learning rediscovered. Science, 247, 1849–1850.
- Bates, E., Thal, D., & Marchman, V. (1991). Symbols and syntax: A Darwinian approach to language development. In N. A. Krasnegor, D. M. Rumbaugh, R. L. Schiefelbusch, & M. Studdert-Kennedy (Eds.), Biological and behavioral determinants of language development (pp. 29–66). Hillsdale, NJ: Erlbaum.
- Bernhardt, B. H., & Stemberger, J. P. (1998). Handbook of phonological development from the perspective of constraint-based nonlinear phonology. San Diego, CA: Academic Press.
- Bernicot, J., & Roux, M. (1998). La structure et l’usage des enonces: Comparison d’enfants uniques et d’enfants second nes [The structure and use of expressions: Comparing only children and second born children]. In J. Bernicot, H. Marcos, C. Day, M. Guidetti, J. Rabain-Jamin, V. Laval, & G. Babelot (Eds.), De l’usage des gestes et des mots chez les enfants [Children’s use of gestures and words] (pp. 157–178). Paris: Armand Colin.
- Bever, T. G., & Chiarello, R. J. (1974). Cerebral dominance in musicians and nonmusicians. Science, 185, 537–539.
- Bickerton, D. (1984). The language bioprogram hypothesis. Behavioral and Brain Sciences, 7, 173–221.
- Bickerton, D. (1988). Creole languages and the bioprogram. In F. J. Newmeyer (Ed.), Linguistics: The Cambridge survey (Vol. II, pp. 268–284). Cambridge, England: Cambridge University Press.
- Birdsong, D. (1999). Introduction: Whys and why nots of the critical period hypothesis for second language acquisition. In D. Birdsong (Ed.), Second language acquisition and the critical period hypothesis (pp. 1–22). Mahwah, NJ: Erlbaum.
- Bloom, L. (1998). Language acquisition in its developmental context. In D. Kuhn & R. S. Siegler (Eds.), Handbook of child psychology: Vol. 2. Cognition, perception, and language (5th ed., pp. 309–370). New York: Wiley.
- Bloom, L., Lifter, K., & Hafitz, J. (1980). Semantics of verbs and the development of verb inflection in child language. Language, 56, 386–412.
- Bornstein, M. H., Haynes, M. O., & Painter, K. M. (1998). Sources of child vocabulary competence: A multivariate model. Journal of Child Language, 25, 367–393.
- Callanan, M. A. (1985). How parents label objects for young children: The role of input in the acquisition of category hierarchies. Child Development, 56, 508–523.
- Carpenter, M., Nagell, K., & Tomasello, M. (1998). Social cognition, joint attention, and communicative competence from 9 to 15 months of age. Monographs of the Society for Research in Child Development, 63(4, Serial No. 255).
- Carstairs-McCarthy, A. (1999). The origins of complex language: An inquiry into the evolutionary beginnings of sentences, syllables, and truth. Oxford, UK: Oxford University Press.
- Chiarello, C. (1991). Interpretation of word meanings by the cerebral hemispheres: One is not enough. In P. J. Schwanenflugel (Ed.), The psychology of word meanings (pp. 251–278). Hillsdale, NJ: Erlbaum.
- Chien, Y.-C. & Wexler, K. (1990). Children’s knowledge of locality conditions in binding as evidence for the modularity of syntax and pragmatics. Language Acquisition, 1, 225–295.
- Chomsky, N. (1965). Aspects of the theory of syntax. Cambridge, MA: MIT Press.
- Chomsky, N. (1982). Discussion of Putnam’s comments. In M. Piattelli-Palamarini (Ed.), Language and learning: The debate between Jean Piaget and Noam Chomsky (pp. 310–324). Cambridge, MA: Harvard University Press.
- Clark, E. V. (1997). Conceptual perspective and lexical choice in acquisition. Cognition, 64, 1–37.
- Conboy, B. T., & Mills, D. (2001, April 19–22). Two languages, one developing brain: Effects of vocabulary size on bilingual toddlers’ event-related potentials to auditory words. Poster presented at the meetings of the Society for Research in Child Development, Minneapolis, MN.
- Crain, S. (1991). Language acquisition in the absence of experience. Behavioral and Brain Sciences, 14, 597–650.
- Curtiss, S. (1977). Genie: A psycholinguistic study of a modern day “wild child.” New York: Academic Press.
- Dale, P. S., Dionne, G., Eley, T. C., & Plomin, R. (2000). Lexical and grammatical development: a behavioral genetic perspective. Journal of Child Language, 27, 619–642.
- deVilliers, J. (1985). Learning how to use verbs: lexical coding and the influence of the input. Journal of Child Language, 12, 587–595.
- Dinnsen, D. A., & O’Connor, K. M. (2000). Typological predictions in developmental phonology. Journal of Child Language, 28, 597–628.
- Dollaghan, C. (1985). Child mets word: “Fast mapping” in preschool children. Journal of Speech and Hearing Research, 28, 449–454.
- Dunn, J., & Shatz, M. (1989). Becoming a conversationalist despite (or because of) having an older sibling. Child Development, 60, 399–410.
- Eley, T. C., Bishop, D. V. M., Dale, P. S., Oliver, B., Petrill, S. A., Price, T. S., Purcell, S., Saudino, K. J., Simonoff, E., Stevenson, J., & Plomin, R. (1999). Genetic and environmental origins of verbal and performance components of cognitive delay in 2-year-olds. Developmental Psychology, 35, 1122–1131.
- Elman, J. L. (1993). Learning and development in neural networks: The importance of starting small. Cognition, 48, 71–99.
- Ferguson, C. A., & Farwell, C. B. (1975). Words and sounds in early language acquisition. Language, 51, 439–491.
- Fodor, J. (1997, May 16). Do we have it in us? Times Literary Supplement, 3–4.
- Fivush, R. (1991). The social construction of personal narratives. Merrill-Palmer Quarterly, 37, 59–81.
- Ganger, J., Pinker, S., Chawla, S., & Baker, A. (2002). Heritability of early milestones of vocabulary and grammar: A twin study. Unpublished manuscript.
- Gathercole, V. C. M. (2002). Monolingual and bilingual acquisition: Learning different treatments of that-trace phenomena in English and Spanish. In D. K. Oller & R. E. Eilers (Eds.), Language and literacy in bilingual children (pp. 220–254). Clevedon, UK: Multilingual Matters.
- Gazzaniga, M. S. (1983). Right hemisphere language following brain bisection: A twenty-year perspective. American Psychologist, 38, 525–537.
- Gilger, J. W. (1996). How can behavioral genetic research help us understand language development and disorders? In Mabel L. Rice (Ed.), Toward a genetics of language (pp. 77–110). Mahwah, NJ: Erlbaum.
- Gillette,J.,Gleitman,H.,Gleitman,L.,&Lederer,A.(1999).Human simulations of vocabulary learning. Cognition, 73, 135–176.
- Gleitman, L. R. (1981). Maturational determinants of language growth. Cognition, 10, 103–114.
- Goldin-Meadow, S. (1997). The resilience of language in humans. In C. Snowdon & M. Hausberger (Eds.), Social influences on vocal development (pp. 293–311). Cambridge, UK: Cambridge University Press.
- Golinkoff, R. M., Mervis, C. B., & Hirsh-Pasek, K. (1994). Early object labels: The case for a developmental lexical principles framework. Journal of Child Language, 21, 125–156.
- Goodman, J., McDonough, L., & Brown, N. (1998). The role of semantic context and memory in the acquisition of novel nouns. Child Development, 69, 1330–1344.
- Gopnik, M., & Crago, M. B. (1991). Familial aggregation of a developmental language disorder. Cognition, 39, 1–50.
- Gould, S. J., & Lewontin, R. C. (1979). The spandrels of San Marco and the Panglossian paradigm: A critique of the adaptationist programme. Proceedings of the Royal Society of London, 205, 581–598.
- Grimshaw, J., & Rosen, S. (1990). Knowledge and obedience: The developmental status of the binding theory. Linguistic Inquiry, 21, 189–222.
- Grodzinsky, Y., & Reinhart, T. (1993). The innateness of binding and development of coreference: A reply to Grimshaw and Rosen. Linguistic Inquiry, 24, 69–103.
- Harris, M., Jones, D., Brookes, S., & Grant, J. (1986). Relations between the non-verbal context of maternal speech and rate of language development. British Journal of Developmental Psychology, 4, 261–268.
- Hart, B., & Risley, T. (1995). Meaningful differences in the everyday experience of young American children. Baltimore: Brookes.
- Hoff, E. (2001). Language Development. Wadsworth: Thompson Learning. California.
- Hoff, E., Laursen, B., & Tardif, T. (2002). Socioeconomic status and parenting. In M. H. Bornstein (Ed.), Handbook of parenting (2nd ed., pp. 231–252). Mahwah, NJ: Erlbaum.
- Hoff, E., & Naigles, L. (2002). How children use input to acquire a lexicon. Child Development, 73, 418–433.
- Hoff-Ginsberg, E. (1985). Some contributions of mothers’speech to their children’s syntactic growth. Journal of Child Language, 12, 367–385.
- Hoff-Ginsberg, E. (1986). Function and structure in maternal speech: Their relation to the child’s development of syntax. Developmental Psychology, 22, 155–163.
- Hoff-Ginsberg, E. (1990). Maternal speech and the child’s development of syntax: A further look. Journal of Child Language, 17, 85–99.
- Hoff-Ginsberg, E. (1994). Influences of mother and child on maternal talkativeness. Discourse Processes, 18, 105–117.
- Hoff-Ginsberg, E. (1998). The relation of birth order and socioeconomic status to children’s language experience and language development. Applied Psycholinguistics, 19, 603–630.
- Hoff-Ginsberg, E. (1999). Formalism or functionalism? Evidence from the study of language development. In M. Darnell, E. Moravscik, M. Noonan, F. Newmeyer, & K. Wheatley (Eds.), Functionalism and formalism in linguistics (pp. 317–340). Amsterdam: John Benjamins.
- Hollich, G. J., Hirsh-Pasek, K., & Golinkoff, R. M. (2000). Breaking the language barrier: An emergentist coalitition model for the origins of word learning. Monographs of the Society for Research in Child Development, 65 (3, Serial No. 262).
- Huttenlocher, J., Haight, W., Bryk, A., Seltzer, M., & Lyons, T. (1991). Early vocabulary growth: Relation to language input and gender. Developmental Psychology, 27, 236–248.
- Jia, G., & Aaronson, D. (1999). Age differences in second language acquisition: The dominant language switch and maintenance hypothesis. In A. Greenhill, H. Littlefield, & C. Tano (Eds.), Proceedings of the 23rd Annual Boston University Conference on Language Development (pp. 301–312). Somerville, MA: Cascadilla Press.
- Jourdan, C. (1991). Pidgins and creoles: The blurring of categories. Annual Review of Anthropology, 20, 187–209.
- Jusczyk, P. W. (1997). The discovery of spoken language. Cambridge, MA: MIT Press.
- Kako, E. (1999). Elements of syntax in the systems of three language-trained animals. Animal Learning and Behavior, 27, 1–14.
- Kelly, M. H., & Martin, S. (1994). Domain-general abilities applied to domain-specific tasks: Sensitivity to probabilities in perception, cognition, and language. In L. Gleitman & B. Landau (Eds.), The acquisition of the lexicon (pp. 105–140). Cambridge, MA: MIT Press.
- Kim, K. H. S., Relkin, N. R., Lee, K.-M., & Hirsch, J. (1997). Distinct cortical areas associated with native and second languages. Nature, 388, 171–174.
- Lieberman, P. (1975). On the origins of language: An introduction to the evolution of human speech. New York: Macmillan.
- Lieven, E. V. M. (1994). Crosslinguistic and crosscultural aspects of language addressed to children. In C. Gallaway & B. J. Richards (Eds.), Input and interaction in language acquisition (pp. 56– 73). Cambridge, UK: Cambridge University Press.
- Macken, M. A., & Ferguson, C. (1983). Cognitive aspects of phonological development: Model, evidence, and issues. In K. E. Nelson (Ed.), Children’s language (Vol. 4, pp. 256–282). Hillsdale, NJ: Erlbaum.
- MacWhinney, B. (1999). The emergence of language. Mahwah, NJ: Erlbaum.
- Maratsos, M. (1998). The acquisition of grammar. In D. Kuhn & R. S. Siegler (Eds.), Handbook of child psychology: Vol. 2. Cognition, perception, and language (5th ed., pp. 421–466). New York: Wiley.
- Marcus, G. F. (1998). Rethinking eliminative connectionism. Cognitive Psychology, 37, 243–282.
- Marcus, G. F. (2001). The algebraic mind. Cambridge, MA: MIT Press.
- Marcus, G. F., Vijayan, S., Bandi Rao, S., & Vishton, P. M. (1999). Rule learning by seven-month-old infants. Science, 283, 77–80.
- Markman, E. M., & Hutchinson, J. E. (1984). Children’s sensitivity to constraints on word meaning: Taxonomic vs. thematic relations. Cognitive Psychology, 16, 1–27.
- Markman, E. M., & Wachtel, G. A. (1988). Children’s use of mutual exclusivity to constrain the meanings of words. Cognitive Psychology, 20, 121–157.
- Mazziotta, J. C., & Metter, E. J. (1988). Brain cerebral metabolic mapping of normal and abnormal language and its acquisition during development. In F. Plum (Ed.), Language, communication, and the brain (pp. 245–266). New York: Raven Press.
- McCabe, A., & Peterson, C. (1991). Getting the story: A longitudinal study of parental styles in eliciting narratives and developing narrative skill. In A. McCabe & C. Peterson (Eds.), Developing narrative structure (pp. 217–253). Hillsdale, NJ: Erlbaum.
- McClelland, J. L., & Plaut, D. C. (1999). Does generalization in infant learning implicate abstract algebra-like rules? Trends in Cognitive Sciences, 3, 166–168.
- Merriman, W. E., & Bowman, L. L. (1989). The mutual exclusivity bias in children’s word learning. Monographs of the Society for Research in Child Development, 54 (3–4, No. 220).
- Mervis, C., & Bertrand, J. (1994). Acquisition of the novel namenameless category principle. Child Development, 65, 1646– 1662.
- Mills, D. L., Coffey-Corina, S., & Neville, H. (1997). Language comprehension and cerebral specialization from 13 to 20 months. Developmental Neuropsychology, 13, 395–445.
- Minami, M., & McCabe, A. (1995). Rice balls and bear hunts: Japanese and North American family narrative patterns. Journal of Child Language, 22, 423–446.
- Morgan, J., & Demuth, K. (1996). Signal to syntax: Bootstrapping from speech to grammar in early acquisition. Hillsdale, NJ: Erlbaum.
- Naigles, L. (1990). Children use syntax to learn verb meanings. Journal of Child Language, 17, 357–374.
- Naigles, L. (1996). The use of multiple frames in verb learning via syntactic bootstrapping. Cognition, 58, 221–251.
- Naigles, L. (1998). Developmental changes in the use of structure in verb learning. In C. Rovee-Collier, L. Lipsitt, & H. Haynes (Eds.), Advances in infancy research (Vol. 12, pp. 298–318). London: Ablex.
- Naigles, L. R., & Hoff-Ginsberg, E. (1998). Why are some verbs learned before other verbs? Effects of input frequency and structure on children’s early verb use. Journal of Child Language, 25, 95–120.
- Nelson, D. A. (1997). Social interaction and sensitive phases for song learning: A critical review. In C. T. Snowdon & M. Hausberger (Eds.), Social influences on vocal development (pp. 7–23). New York: Cambridge University Press.
- Nelson, K. (1988). Constraints on word learning? Cognitive Development, 3, 221–246.
- Neville, H. J., Nicol, J. L., Barss, A., Forster, K. I., & Garrett, M. (1991). Syntactically based sentence processing classes: Evidence from event-related brain potentials. Journal of Cognitive Neuroscience, 3, 151–165.
- Newport, E. L., & Aslin, R. N. (2000). Innately constrained learning: Blending old and new approaches to language acquisition. In S. C. Howell, S. A. Fish, & T. Keith-Lucas (Eds.), Proceedings of the 24th Boston University Conference on Language Development (pp. 1–21). Somerville, MA: Cascadilla Press.
- Newport, E. L., Gleitman, H., & Gleitman, L. R. (1977). Mother, I’d rather do it myself: Some effects and non-effects of maternal speech style. In C. E. Snow & C. A. Ferguson (Eds.), Talking to children: Language input and acquisition (pp. 109–150). Cambridge, UK: Cambridge University Press.
- Ninio, A., & Snow, C. E. (1999). The development of pragmatics: Learning to use language appropriately. In W. C. Ritchie & T. K. Bhatia (Eds.), Handbook of child language acquisition (pp. 347–386). New York: Academic Press.
- Ochs, E., & Schieffelin, B. B. (1979). Developmental pragmatics. New York: Academic Press.
- Pearson, B. Z., Fernandez, S. C., Lewedeg, V., & Oller, D. K. (1997). The relation of input factors to lexical learning by bilingual infants. Applied Psycholinguistics, 18, 41–58.
- Perani, D., Paulesu, E., Galles, N. S., Dupous, E., Dehaene, S., Bettinardi, V., Cappa, S. F., Fazio, F., & Mehler, J. (1998). The bilingual brain: Proficiency and age of acquisition of the second language. Brain, 121, 1841–1852.
- Pinker, S. (1984). Language learnability and language development. Cambridge, MA: Harvard University Press.
- Pinker, S. (1994). The language instinct: How the mind creates language. New York: Morrow.
- Pinker, S. (1999). Words and rules. New York: Basic Books.
- Pinker, S., & Bloom, P. (1990). Natural language and natural selection. Behavioral and Brain Sciences, 13, 707–784.
- Plaut, D., & Kello, C. (1999). The emergence of phonology from the interplay of speech comprehension and production: Adistributed connectionist approach. In B. MacWhinney (Ed.), The emergence of language (pp. 381–416). Hillsdale, NJ: Erlbaum.
- Plunkett, K. (Ed.). (1998). Language acquisition and connectionism. East Sussex, UK: Psychology Press.
- Prasada, S., & Choy, J. (1998, November). The role of syntactic structure in the interpretation of proper nouns. Paper presented at the 23rd Annual Boston University Conference on Language Development, Boston.
- Premack, D. (1986). Gavagai! Or the future history of the animal language controversy. Cambridge, MA: MIT Press.
- Reese, E., & Fivush, R. (1993). Parental styles of talking about the past. Developmental Psychology, 29, 596–606.
- Reznick, J. S., Corley, R., & Robinson, J. (1997). A longitudinal twin study of intelligence in the second year. Monographs of the Society for Research in Child Development, 62 (1, Serial No. 249).
- Saffran, J. R., Aslin, R. N., & Newport, E. L. (1996). Statistical learning by 8-month-old infants. Science, 274, 1926–1928.
- Schieffelin, B. B., & Ochs, E. (Eds.). (1986). Language socialization across cultures. Cambridge, UK: Cambridge University Press.
- Senghas, A. (1995). The development of Nicaraguan Sign Language via the language acquisition process. In D. MacLaughlin & S. McEwen (Eds.), Proceedings of the 19th Boston University Conference on Language Development (pp. 543–552). Somerville, MA: Cascadilla Press.
- Senghas,A. (2000). The development of early spatial morphology in NicaraguanSignLanguage.InS.C.Howe,S.A.Fish,&T.KeithLucas (Eds.), Proceedings of the 24th Boston University Conference on Language Development (pp. 696–707). Somerville, MA: Cascadilla Press.
- Shatz, M. (1992). A forward or backward step in the search for an adequate theory of language acquisition? Social Development, 1,
- Shatz, M. (1994). A toddler’s life: Becoming a person. New York: Oxford University Press.
- Shatz, M., Hoff-Ginsberg, E., & MacIver, D. (1989). Induction and the acquisition of English auxiliaries: the effects of differentially enriched input. Journal of Child Language, 16, 121–140.
- Siskind, J. M. (1996). A computational study of cross-situational techniques for learning word-to-meaning mappings. In M. R. Brent (Ed.), Computational approaches to language acquisition (pp. 39–91). Cambridge, MA: MIT Press.
- Slobin, D. I., Gerhardt, J., Kyratzis, & Guo, J. (Eds.). (1996). Social interaction, social context, and language: Essays in honor of Susan Ervin-Tripp. Mahwah, NJ: Erlbaum.
- Snow, C. E., & Hoefnagel-Hohle, M. (1978). The critical period for language acquisition: Evidence from second language learning. Child Development, 49, 1114–1128.
- Springer, A., & Deutsch, G. (1981). Left brain, right brain. San Francisco: W. H. Freeman.
- Stiles, J., Bates, E. A., Thal, D., Trauner, D., & Reilly, J. (1998). Linguistic, cognitive, and affective development in children with pre- and perinatal focal brain injury: A ten-year overview from the San Diego longitudinal project. In C. Rovee-Collier, L. Lipsitt, & H. Hayne (Eds.), Advances in Infancy Research (Vol. 12, pp. 131–164). Stamford, CT: Ablex.
- Stromswold, K. (1998). The genetics of spoken language disorders. Human Biology, 70, 297–324.
- Stromswold, K. (2001). The heritability of language: A review and metaanalysis of twin, adoption and linkage studies. Language, 77, 647–723.
- Tamis-LeMonda, C. S., Bornstein, M. H., Kahan-Kalman, R., Baumwell, L., & Cyphers, L. (1998). Predicting variation in the timing of linguistic milestones in the second year: An eventshistory approach. Journal of Child Language, 25, 675–700.
- Theakston, A., Lieven, E., Pine, J., & Rowland, C. (2001). The role of performance limitations in the acquisition of verb-argument structure: an alternative account. Journal of Child Language, 28, 127–152.
- Tomasello, M. (1992). First verbs: A case study of early grammatical development. Cambridge: Cambridge University Press.
- Tomasello, M. (1992). The social bases of language acquisition. Social Development, 1, 67–87.
- Tomasello, M. (1994). Can an ape understand a sentence? A review of Language comprehension in ape and child by E. S. SavageRumbaugh et al. Language & Communication, 14, 377–390.
- Tomasello, M. (2000). Do young children have adult syntactic competence? Cognition, 74, 209–253.
- Tomasello, M., Call, J., Nagell, K., Olguin, R., & Carpenter, M. (1994). The learning and use of gestural signals by young chimpanzees: A trans-generational study. Primates, 35, 137–154.
- Tomasello, M., & Farrar, M. J. (1986). Joint attention and early language. Child Development, 57, 1454–1463.
- Tomasello, M., Kruger, A., & Ratner, H. (1993). Cultural learning. Behavioral and Brain Sciences, 16, 495–552.
- Tomasello, M., & Todd, J. (1983). Joint attention and lexical acquisition style. First Language, 4, 197–212.
- Valian, V. (1991). Syntactic subjects in the early speech of American and Italian children. Cognition, 40, 21–81.
- Vear, D., Naigles, L., Hoff, E., & Ramos, E. (2001, December, 5–8). An investigation of the syntactic flexibility within young children’s early verb development: Evidence from a cross-sectional diary study. Paper presented to the Early Lexicon Acquisition Conference, Lyon, France.
- Weber-Fox, C. M., & Neville, H. J. (1996). Maturational constraints on functional specializations for language processing: ERP and behavioral evidence in bilingual speakers. Journal of Cognitive Neuroscience, 8, 231–256.
- Wexler, K. (1999). Maturation and growth of grammar. In W. C. Ritchie & T. K. Bhatia (Eds.), Handbook of child language acquisition (pp. 55–111). San Diego, CA: Academic Press.
- Weylman, S. T., Brownell, H. H., & Gardner, H. (1988). “It’s what you mean, not what you say”: Pragmatic language use in braindamaged patients. In F. Plum (Ed.), Language, communication, and the brain (pp. 229–244). New York: Raven Press.
- Woodward, A. L., & Markman, E. M. (1998). Early word learning. In D. Kuhn & R. S. Siegler (Eds.), Handbook of child psychology: 2. Cognition, perception, and language (5th ed., pp. 371–420). New York: Wiley.
- Woodward, A. L., Markman, E. M., & Fitzsimmons, C. M. (1994). Rapid word learning in 13- and 18-month-olds. Developmental Psychology, 30, 553–556.